We have obtained pedigree data, blood and archived tissue from members of 98 families containing three or more confirmed cases of prostate cancer. We have typed these families for 437 autosomal markers and have analyzed the data for linkage. Although no single chromosomal region met the genome-wide criteria for statistically significant excess allele sharing, our strongest signal on the distal end of chromosome 19p has been replicated independently by a study of multiple-case prostate cancer families in Sweden (Wiklund et al. 2003). In this competing renewal application we request funds to test the hypothesis that region 19p13.3 harbors a prostate cancer susceptibility gene. Our objectives are to: 1) narrow the region by tripling the number of markers and perform linkage analysis to exclude subregions with low lod scores, using statistical methods that accommodate both individual-specific and family-specific covariates that may account for genetic heterogeneity across families; 2a) rank the 92 known genes and 61 transcripts of unknown genes in the (narrowed) region with respect to their potential involvement in prostate cancer, and 2b) identify polymorphisms in the more promising genes; 3a) investigate associations between prostate cancer risk and the variant alleles in the identified polymorphisms by genotyping 750 African-American and 750 Caucasian case-control pairs in a casecontrol study nested within the Hawaii/Los Angeles multiethnic cohort (MEC); 3b) when warranted, investigate the functional significance of these variants. The NCI-funded International Consortium for Prostate Cancer Genetics (ICPCG) is a resource of more than 1,500 multiple-case prostate cancer families. As members of the ICPCG, our long-term aim is to pursue with this group any promising leads identified in this project.